[
https://issues.apache.org/jira/browse/QPID-4086?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Robbie Gemmell updated QPID-4086:
---------------------------------
Assignee: (was: Robbie Gemmell)
Status: Open (was: Ready To Review)
Looking over the patch several times and almost being ready to apply it, I have
come to think it resolves this specific case but potentially creates various
other similar issues. Reversing the locks in the close() method would fix this
instance, but because the Dispatcher thread acquires the message delivery lock
and then application code is able to use that thread to perform work in
onMessage, it would then be possible for the work undertaken to acquire the
other lock in a variety of locations and end up creating new issues.
Given this is a long standing issue, the proximity to the next release, and the
fact this particular scenario (failover attempts being exhausted at the exact
same time an application-initiated connection close occurs) isn't that likely
in typical use I think we need to defer making immediate changes to this area
and look at a longer term solution.
I have updated the specific tests that inadvertently hit the issue (via
QPID-4080) sporadically to stop them doing so, and we can add a more targeted
test when any work is done on this JIRA.
> Java Client 0-9.x deadlocks on concurrent connection close and session close
> from failover thread
> -------------------------------------------------------------------------------------------------
>
> Key: QPID-4086
> URL: https://issues.apache.org/jira/browse/QPID-4086
> Project: Qpid
> Issue Type: Bug
> Components: Java Client
> Affects Versions: 0.16, 0.17
> Reporter: Alex Rudyy
> Attachments: 0001-QPID-4086-Fix-deadlock.patch, deadlock.txt
>
>
> Java Client 0-8/0-9.x deadlocks on concurrent connection close and session
> close from failover thread
> {noformat}
> Java stack information for the threads listed above:
> ===================================================
> "Failover":
> at org.apache.qpid.client.AMQSession.closed(AMQSession.java:822)
> - waiting to lock <0x00000000e03c3160> (a java.lang.Object)
> at
> org.apache.qpid.client.AMQConnection.closeAllSessions(AMQConnection.java:954)
> at
> org.apache.qpid.client.AMQConnection.exceptionReceived(AMQConnection.java:1298)
> - locked <0x00000000e03c71b8> (a java.lang.Object)
> at
> org.apache.qpid.client.failover.FailoverHandler.run(FailoverHandler.java:187)
> - locked <0x00000000e03c71b8> (a java.lang.Object)
> at java.lang.Thread.run(Thread.java:662)
> "main":
> at org.apache.qpid.client.AMQConnection.doClose(AMQConnection.java:857)
> - waiting to lock <0x00000000e03c71b8> (a java.lang.Object)
> - locked <0x00000000e03c71c8> (a java.lang.Object)
> at org.apache.qpid.client.AMQConnection.doClose(AMQConnection.java:848)
> - locked <0x00000000e03c3160> (a java.lang.Object)
> - locked <0x00000000e03c71c8> (a java.lang.Object)
> at org.apache.qpid.client.AMQConnection.close(AMQConnection.java:832)
> at org.apache.qpid.client.AMQConnection.close(AMQConnection.java:823)
> at org.apache.qpid.client.AMQConnection.close(AMQConnection.java:818)
> at
> org.apache.qpid.test.utils.QpidBrokerTestCase.tearDown(QpidBrokerTestCase.java:1177)
> at junit.framework.TestCase.runBare(TestCase.java:130)
> at
> org.apache.qpid.test.utils.QpidBrokerTestCase.runBare(QpidBrokerTestCase.java:239)
> at junit.framework.TestResult$1.protect(TestResult.java:106)
> at junit.framework.TestResult.runProtected(TestResult.java:124)
> at junit.framework.TestResult.run(TestResult.java:109)
> at junit.framework.TestCase.run(TestCase.java:118)
> at org.apache.qpid.test.utils.QpidTestCase.run(QpidTestCase.java:137)
> at junit.framework.TestSuite.runTest(TestSuite.java:208)
> at junit.framework.TestSuite.run(TestSuite.java:203)
> at
> org.apache.tools.ant.taskdefs.optional.junit.JUnitTestRunner.run(JUnitTestRunner.java:520)
> at
> org.apache.tools.ant.taskdefs.optional.junit.JUnitTestRunner.launch(JUnitTestRunner.java:1060)
> at
> org.apache.tools.ant.taskdefs.optional.junit.JUnitTestRunner.main(JUnitTestRunner.java:911)
> Found 1 deadlock.
> {noformat}
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
